In patients with heart disease, changes in left ventricular filling pressures produce alterations in the Doppler transmitral flow velocity and isovolumetric relaxation time. This investigation explored the hypothesis that combining isovolumetric relaxation time with measurements derived from the transmitral flow velocity can be used to estimate left ventricular end-diastolic pressure. Simultaneous Doppler and left ventricular pressure recordings were obtained in 33 patients (24 men with a mean age of 58 +/- 11 years) and an ejection fraction ranging from 15% to 74% (mean 55 +/- 15%). The following Doppler measurements correlated significantly with left ventricular end-diastolic pressure (range 4 to 36 mm Hg): isovolumetric relaxation time (IVRT; r = -0.73), atrial filling fraction (AFF; r = -0.66), deceleration time (DT; r = -0.59), ratio of early transmitral flow velocity to atrial flow velocity (E/A ratio; r = -0.53) and time from termination of mitral flow to the electrocardiographic R wave (MAR; r = 0.37). Combining these measurements into a multilinear regression equation provided a more accurate estimate of end-diastolic pressure (LVEDP; r = 0.80; SEE = 7.4). The equation LVEDP = 46 -0.22 IVRT -0.10 AFF -0.03 DT -(2 divided by E/A) + 0.05 MAR was tested prospectively in 26 additional patients (mean age 55 +/- 11 years; ejection fraction 41 +/- 23%) with simultaneous Doppler and hemodynamic recordings but with the two measurements made independently, in blinded fashion, by additional observers. Estimated and measured end-diastolic pressures correlated well with each other (r = 0.86).(ABSTRACT TRUNCATED AT 250 WORDS)